Original Article
HIP1–ALK, A Novel ALK Fusion Variant that Responds to Crizotinib Douglas D. Fang, PhD,* Bin Zhang, PhD,* Qingyang Gu, PhD,* Maruja Lira, BS,‡ Qiang Xu, PhD,† Hongye Sun, PhD,† Maoxiang Qian, PhD,† Weiqi Sheng, MD, PhD,§ Mark Ozeck, BS,‡ Zhenxiong Wang, PhD,‡ Cathy Zhang, MS,‡ Xinsheng Chen, MS,* Kevin X. Chen, PhD,* Jian Li, PhD,* Shu-Hui Chen, PhD,* James Christensen, PhD,‡ Mao Mao, MD, PhD,‡ and Chi-Chung Chan, PhD* Introduction: The aim of this study was to identify anaplastic lymphoma kinase (ALK) rearrangements in lung cancer patient-derived xenograft (PDX) models and to explore their responses to crizotinib. Methods: Screening of 99 lung cancer PDX models by the NanoString ALK fusion assay identified two ALK-rearranged non– small-cell lung cancer (NSCLC) tumors, including one harboring a previously known echinoderm microtubule-associated protein-like 4 (EML4)–ALK fusion and another containing an unknown ALK fusion variant. Expression array, RNA-Seq, reverse transcription polymerase chain reaction, and direct sequencing were then conducted to confirm the rearrangements and to identify the novel fusion partner in the xenograft and/or the primary patient tumor. Finally, pharmacological studies were performed in PDX models to evaluate their responses to ALK inhibitor crizotinib. Results: Two ALK-rearranged NSCLC PDX models were identified: one carried a well-known EML4–ALK variant 3a/b and the other harbored a novel huntingtin interacting protein 1 (HIP1)–ALK fusion gene. Exon 28 of the HIP1 gene located on chromosome 7 was fused to exon 20 of the ALK gene located on chromosome 2. Both cases were clinically diagnosed as squamous cell carcinoma. Compared with the other lung cancer PDX models, both ALK-rearranged models displayed elevated ALK mRNA expression. Furthermore, in vivo efficacy studies demonstrated that, similar to the EML4–ALK-positive model, the HIP1–ALKcontaining PDX model was sensitive to treatment with crizotinib. Conclusions: Discovery of HIP1 as a fusion partner of ALK in NSCLC is a novel finding. In addition, the HIP1–ALK-rearranged *Discovery Services, †Genomics Center, WuXi AppTec Co., Ltd., Shanghai, China; ‡Pfizer Oncology Research Unit, San Diego, California; and §Department of Pathology, Shanghai Cancer Center, Fudan University, Shanghai, China. Drs. Zhang, and Gu contributed equally to this work. Drs. Mao and Chan contributed equally as senior authors. Disclosure: Several authors, as indicated above, are currently employees of Pfizer, at which crizotinib was invented, developed, and manufactured. The other authors declare no conflict of interest. Address for correspondence: Chi-Chung Chan, PhD, WuXi AppTec Co., Ltd., 288 Fute Zhong Road, Waigaoqiao Free Trade Zone, Shanghai, China 200131. E-mail: [email protected]; or Mao Mao, MD, PhD, Pfizer Oncology Unit, 10777 Science Center Drive, San Diego, CA 92121. E-mail: [email protected]. Copyright © 2014 by the International Association for the Study of Lung Cancer ISSN: 1556-0864/14/0903-0285
tumor is sensitive to treatment with crizotinib in vivo, implicating HIP1–ALK as an oncogenic driver of lung tumorigenesis. Collectively, our results indicate that HIP1–ALK-positive NSCLC may benefit from clinical applications of crizotinib. Key Words: Huntingtin interacting protein 1, anaplastic lymphoma kinase, HIP1–ALK, Echinoderm microtubule-associated proteinlike 4, EML4–ALK, Non–small-cell lung cancer, Crizotinib (PF02341066), Patient-derived xenograft (PDX) tumor model. (J Thorac Oncol. 2014;9: 285–294)
T
yrosine kinases have been widely recognized as attractive targets for molecularly targeted cancer therapy. Over the last decade a greater understanding of the molecular characteristics of non–small-cell lung cancer (NSCLC) has led to the development of effective targeted therapies for personalized medicine in a subset of patients. Targeting activating epidermal growth factor receptor (EGFR) mutations was the first breakthrough. Activation by mutation in EGFR in NSCLC correlated with clinical response to EGFR tyrosine kinase inhibitors.1,2 Afterward, in 2007, echinoderm microtubule-associated protein-like 4 (EML4)–ALK fusion-type tyrosine kinase was identified as an oncoprotein in 4% to 5% of NSCLC cases.3 Clinical trials were soon initiated to investigate the role of crizotinib (PF-02341066), an oral selective inhibitor of ALK and hepatocyte growth factor receptor (c-MET) kinases,4 for the treatment of EML4–ALK-positive NSCLC.5 On the basis of response rates reported in phase 1 and 2 clinical trials, crizotinib received accelerated approval by U.S. Food and Drug Administration in August 2011 for the treatment of patients with locally advanced or metastatic NSCLC that is ALK-positive as detected by the ALK fluorescence in situ hybridization test. EML4–ALK is generated as a result of a small inversion within the short arm of human chromosome 2.3,6,7 EML4–ALK undergoes constitutive dimerization through interaction between the coiled-coil domain within the EML4 region of each monomer, thereby activating ALK and generating oncogenic activity. Besides EML4, other fusion partners of ALK have been reported in a variety of cancer types (primarily in hematopoietic malignancies), including nucleophosmin (NPM),8,9 TRK-fused gene (TFG),10 clathrin heavy chain (Hc) (CLTC),11
Journal of Thoracic Oncology ® • Volume 9, Number 3, March 2014
285
Journal of Thoracic Oncology ® • Volume 9, Number 3, March 2014
Fang et al
5-aminoimidazole-4-carboxamide ribonucleotide formyltransferase/IMP cyclohydrolase (ATIC),12,13 cysteinyl-tRNA synthetase (CARS),14 moesin (MSN),15 tropomyosin 3 (TPM3),16,17 kinesin family member 5B (KIF5B),18 kinesin light chain 1 (KLC1),19 and tyrosine-protein phosphatase non-receptor type 3 (PTPN3).20 To date, only EML4–ALK, TFG–ALK, KIF5B–ALK, KLC1–ALK, and ALK–PTPN3 have been identified in lung cancer.21 Although the biological and clinical significance of these variants requires further investigation, it seems that not all of the above mentioned ALK rearrangements respond to crizotinib. For example, an ALK– PTPN3 variant may not respond to crizotinib, as the ALK kinase domain is absent.21 In addition, other, not-yet-characterized fusions may also exist in solid tumors, including lung cancer.6 We report here a novel ALK fusion associated with NSCLC involving the HIP1 gene and demonstrate the sensitivity of a HIP1–ALK-rearranged tumor to treatment with the ALK inhibitor crizotinib.
MATERIALS AND METHODS Establishment of NSCLC PDX Models In compliance with the protocol approved by the Institutional Review Board of Cancer Institute and Hospital of Chinese Academy of Medical Sciences and with the subject’s informed consent, primary tumor tissues were collected for PDX establishment. In brief, surgically resected primary tumor tissues from patients (designated as PA) were collected, trimmed, cut into 20- to 30-mm3 fragments and immediately implanted subcutaneously by using an 18-gauge trocar in the fore and/or hind bilateral flanks of 6- to 8-week-old female BALB/c nude mice (Shanghai SLAC Laboratory Animal Co., Ltd., Shanghai, China). Once the first generation of xenografts (designed as P0) was established and tumor sizes reached 500 to 800 mm3, serial implantations in BALB/c nude mice were performed to expand the xenograft tumors (i.e., P1, P2, P3, and beyond). Tumor size was measured periodically using a digital caliper (Cal Pro, Sylvac, Switzerland). Tumor volume was calculated as 0.5 × length × width.2 All procedures and protocols were approved by the Institutional Animal Care and Use Committee of WuXi AppTec.
Histology Pieces of patient samples or PDX tissues at each passage were fixed in 10% buffered formalin for 24 to 48 hours and then paraffin-embedded. Five μm sections were cut and stained with hematoxylin and eosin. Histopathology was reviewed by a pathologist (WS).
Total RNA Isolation Total RNA was isolated from the primary patient tumors and xenografts at P3 (LU-01-0015) or P2 (LU-01-0319) mouse models by using an RNeasy protect Mini Kit (Qiagen, Valencia, CA).
NanoString Assay for ALK Fusions The method for the detection of ALK fusions and expressions using NanoString’s nCounter technology (NanoString Technologies, Inc., Seattle, WA) has been described previously.22 In brief, after incubation of total RNA with nCounter probe sets, the samples were transferred to the nCounter Prep Station (NanoString Technologies, Inc.), where excess probes were
286
removed and probe/target complexes were aligned and immobilized in the nCounter Cartridge (NanoString Technologies, Inc.). Cartridges containing immobilized and aligned reporter complex were subsequently imaged on an nCounter Digital Analyzer (NanoString Technologies, Inc.) set at 1155 fields of view. Reporter counts were collected using NanoString’s nSolver analysis software version 1. The raw data were normalized and analyzed in two steps as previously described.22 A single-tube, multiplexed assay system was designed to simultaneously detect ALK fusion transcripts and measure the expression levels of several ALK exons flanking the fusion breakpoint. For fusion detection, ALK fusion partner EML4-, KIF5B-, or TFG-specific 5′ capture probes and ALK-specific 3′ reporter probes were designed to hybridize to ALK fusion partner and ALK at the fusion junction, respectively. The capture probes were able to detect all major variants of EML4-ALK fusions (variants 1, 2, 3a, 5, and “V5”21,23) and non–EML4–ALK fusions, including KIF5B–ALK and TFG– ALK. For ALK gene expression, probe sets across the entire ALK transcript, including four probe sets designated as ALK 5 ′-1 to 5 ′-4 located upstream of the intron 19 fusion breakpoint and four probe sets ALK 3 ′-1 to 3 ′-4 located downstream of the fusion breakpoint, were designed to measure the expression levels of ALK exons flanking the fusion breakpoint. For fusion detection, a reporter count of 60 was designated as the background threshold level, and the fusion was called present if the normalized reporter count was higher than background threshold. For ALK gene expression, the ALK 3′ overexpression score (i.e., ALK3 ′/5′ ratio) was defined as follows: ALK3′/5′ = E3/max(A5, B), where E3 is the geometric mean of ALK 3′ probe expression, A5 is the average of the ALK 5′ probe expression, and B is the background threshold. The fusion was called present if the ALK 3′ overexpression score was greater than two.
Microarray for ALK Expression Total RNA was amplified and fragmented using a GeneChip 3′ IVT Expression Kit (Affymetrix, Santa Clara, CA). Then the samples were hybridized onto a GeneChip PrimeView Human Gene Expression Array or GeneChip Human Genome U133 Plus 2.0 Array (Affymetrix) and scanned on a Affymetrix GeneChip Scanner 3000 7G (Affymetrix). Data were subject to in-house bioinformatics analysis. In brief, the raw CEL data of the PrimeView and U133 Plus 2.0 Arrays were processed on an Expression Console (version 1.1, Affymetrix). Signal intensities of both arrays were normalized by the Robust Multiarray Average normalization approach.
Paired-End RNA-Seq Analysis RNA-Seq library was generated using the Illumina TruSeq RNA Sample Prep Kit v2 (Illumina, San Diego, CA) and sequencing was performed on the Illumina HiSeq 2000 platform (Illumina). Human sequence reads were isolated from the mixture of reads arising from the mouse host and reads arising from xenograft tumors by using a specific tool Xenome.24 The purified reads were then mapped to the Ensembl GRCh37.62 B (hg19) reference genome using an RNA sequence aligner TopHat 2,25 which can align reads across splice junctions with or without relying on gene annotation. On the basis of the
Copyright © 2014 by the International Association for the Study of Lung Cancer
Journal of Thoracic Oncology ® • Volume 9, Number 3, March 2014
aligned result, transcript assembly and abundance estimation were performed with the software Cufflinks, a well-established transcript assembler,26 and fusion transcripts were discovered with the enhanced TopHat-Fusion module.27
Reverse Transcription Polymerase Chain Reaction and Direct Sequencing
Total RNA (1 μg) was reversely transcribed into single-stranded cDNAs using High Capacity cDNA Reverse Transcription Kits (Life Technologies, Carlsbad, CA) following the manufacturer’s instructions. One microliter of cDNA was used for a subsequent 25-μl polymerase chain reaction (PCR) amplification using a Premix PrimeSTAR HS PCR amplification kit (Takara Bio, Otsu, Shiga, Japan). To detect fusion transcript, we designed forward- and reverseprimer combinations targeting the sequences flanking the breakpoint detected by RNA-Seq. For HIP1–ALK fusion, the forward primer (5′-CGGCCAGAATAGAGGGTACA-3′) targeting the 5′ partner gene HIP1 and reverse primer (5′CACCTGGCCTTCATACACCT-3′) targeting the 3′ partner ALK were used. For EML4–ALK fusion, the previously described forward primer (5′-TACCAGTGCTGTCTCAATTGCAGG-3 ′) and reverse primer (5′-TCTTGCCAGCAAAGCAGTAGTTGG-3′) were used to detect EML4–ALK variant 3a/3b28. The PCR products were sequenced by conventional direct sequencing (BioSune Biotechnology (Shanghai) Co., Ltd., Shanghai, China) by using an ABI 3730 DNA Sequencer (Applied Biosystems, Foster City, CA). The results were compared with the reference sequence to validate the fusion transcripts.
In Vivo Efficacy Study Crizotinib [(R)-3-[1-(2,6-dichloro-3-fluoro-phenyl)ethoxy]-5-(1-piperidin4-yl-1H-pyrazol-4-yl)-pyridin-2-ylamine] was synthesized by WuXi AppTec (Shanghai, China) and formulated in distilled water for in vivo study. Injectable cisplatin for clinical use (10 mg/vial) was purchased from Qilu Pharmaceutical Co., Ltd (Jinan, Shandong, China). Tumor tissues were cut into small fragments of approximately 30 mm3 under sterile conditions. BALB/c nude mice were implanted subcutaneously with a tumor fragment by using a trocar. When the average tumor size reached 150 to 200 mm3, tumor size–matched mice were randomly assigned to four groups with six or seven mice in each group. The tumor-bearing mice were given ALK inhibitor crizotinib at 25 and 50 mg/kg (by mouth daily), cisplatin at 3 mg/kg (by intraperitoneal, every week), or vehicle alone for 3 weeks. Tumor volumes and body weights were measured using calipers twice a week. The difference in tumor volumes between treatment groups was analyzed for significance using one-way analysis of variance followed by Dunnett’s test. The p value less than 0.05 was considered to be statistically significant.
RESULTS In the past 2 years, we established more than 150 transplantable lung cancer PDX models from Chinese lung cancer patients. The first 99 of the 150 models, consisting of 96 with NSCLC, were used in this study, and their clinical information is summarized in Supplementary Table 1 (Supplemental Digital Content 1, http://links.lww.com/JTO/A526).
HIP1–ALK, a ALK Fusion Variant that Responds to Crizotinib
Detection of ALK Rearrangement in Lung Cancer PDX Models by Nanostring Fusion Assay To identify ALK rearrangements, RNA samples isolated from xenograft tissues of 99 models were analyzed by a combination of ALK 3′ overexpression- and ALK fusion-specific assays using NanoString’s nCounter technology. Because ALK is generally not expressed in normal adult tissues, ALK 3′ overexpression in lung cancer was indicative of an ALK rearrangement. In the ALK 3′ overexpression assay, we found that two NSCLC models, LU-01-0015 and LU-01-0319, exhibited ALK 3′ overexpression as shown in Figure 1A with an ALK 3′ overexpression score (ALK3 ′/5′ ratio) of 6.19 and 7.28, respectively (Fig. 1B). These results show that ALK is overexpressed in both NSCLC models, implicating ALK rearrangements. ALK fusion partner–specific 5′ capture probes and ALKspecific 3′ reporter probes were designed to detect the presence of preselected ALK fusion partners as previously described.22 Most ALK fusion variants share the same 3′ portion of ALK starting with exon 20. Thus, a common reporter probe (designated as ALK exon 20), paired with capture probes targeting different ALK fusion partners and variants, was able to detect the preselected fusion transcripts containing ALK exon 20 sequences, including all major variants of EML4–ALK fusions (variants 1, 2, 3a, 5, and V5) and KIF5B–ALK and TFG–ALK rearrangements. A reporter count of 60 was designated as the background threshold level, and the fusion was called present if the normalized reporter count was higher than background threshold. Among two models, LU-01-0319 showed higher reporter counts than the fusion reporter threshold, suggesting that it harbored the preselected fusion (Fig. 1C). Conversely, the reporter counts detected by the ALK exon 20 reporter were below the threshold in LU-01-0015. The data indicate that LU-01-0015 likely harbored a novel ALK fusion variant, which was not detectable by using the probe sets specific for known fusion genes indicated above.
Confirmation of ALK Overexpression in LU01-0015 and LU-01-0319 by Microarray The gene expression profiles of 29 lung cancer PDX models, including LU-01-0015 and LU-01-0319, were analyzed by a GeneChip PrimeView Human Gene Expression Array (Affymetrix; 7 models) or GeneChip Human Genome U133 Plus 2.0 Array (Affymetrix; 22 models). Analyses of ALK expression revealed that ALK mRNA levels in both LU-010015 and LU-01-0319 were markedly higher than in the other PDX models (Fig. 1D). These results provide further evidence for the presence of ALK rearrangements in both models.
Identification of HIP1–ALK by RNA-Seq To identify ALK fusion variants in LU-01-0015 and LU-01-0319, we used massively parallel, paired-end sequencing of expressed transcripts (RNA-Seq) to detect gene fusions as described in Materials and Methods. The RNA-Seq produced approximately 80.9 million and 105.5 million paired reads (100 base pair/read) for LU-01-0015 and LU-01-0319, respectively. Data analyses revealed that the ALK fusion gene in LU-010319 was a known EML4–ALK variant 3a/b, which was created by the breakage and intrachromosomal rearrangement of
Copyright © 2014 by the International Association for the Study of Lung Cancer
287
Fang et al
Journal of Thoracic Oncology ® • Volume 9, Number 3, March 2014
FIGURE 1. Detection of ALK rearrangements and overexpression of the ALK gene in two NSCLC PDX models. A, Quantification of ALK 5′ and 3′ expression levels using the NanoString ALK fusion assay. B, ALK 3′ overexpression in both models as measured by ALK 3′/5′ score in the fusion assay. A background threshold of twofold was denoted by a horizontal dash line (B). C, ALK exon 20 reporter counts detected in NanoString assay. A background threshold of 60 reporter counts was denoted by a horizontal dash line (B). D, Expression levels of ALK in NSCLC PDX models determined by a GeneChip PrimeView human gene expression array or a GeneChip Human Genome U133 Plus 2.0 array in a panel of 22 and seven models, respectively. ALK, anaplastic lymphoma kinase; PDX, patient–derived xenograft; NSCLC, non–small-cell lung cancer.
chromosome 2 (Fig. 2A and B).28 Of special interest is the identification of huntingtin interacting protein 1 (HIP1) as a novel fusion partner of ALK in the second model LU-01-0015 (Fig. 2A and B). HIP1–ALK rearrangement was formed by breakage and rejoining of chromosomes 7 and 2. Sixteen spanning reads and 2 spanning mate pairs were aligned to the cDNA breakpoints of HIP1 and ALK genes, which revealed a fusion product formed by exon 28 of HIP1 fused to exon 20 of ALK (Fig. 2B).
288
Validation of ALK Fusions in Both Primary Tumors and Xenografts by RT-PCR and Direct Sequencing To further confirm the fusion transcripts, RT-PCR was conducted by using forward- and reverse-primer combinations pairing the sequences flanking the breakpoint detected by RNA-Seq. Subsequent sequencing of cDNA confirmed the presence of HIP1–ALK rearrangement in both the
Copyright © 2014 by the International Association for the Study of Lung Cancer
Journal of Thoracic Oncology ® • Volume 9, Number 3, March 2014
HIP1–ALK, a ALK Fusion Variant that Responds to Crizotinib
FIGURE 2. Identification of a novel HIP1–ALK fusion by RNA-Seq. A, Circos plots depicting ALK rearrangements in LU-01-0015 and LU-01-0319. Chromosomes were drawn to scale around the rim of the circle, and data are plotted on these coordinates. The blue arc and red arc represent intrachromosomal and interchromosomal translocation, respectively. B, Representative aligned reads were shown to map across the exon–exon fusion junction between HIP1 or EML4 and exon 20 of ALK in LU-010015 and LU-01-0319, respectively, as illustrated in the schematic diagram. In LU-01-0319, the split-reads are shown aligning on the breakpoints of EML4–ALK variant 3a (A) and 3b (B). ALK, anaplastic lymphoma kinase; HIP1, huntingtin interacting protein 1; EML4, echinoderm microtubule-associated protein-like 4.
primary tumor and xenograft of LU-01-0015 (Fig. 3A and B, left panels). In the predicted fusion protein, the majority of N-terminal of HIP1 encompassing the AP180 N-terminal homology (ANTH), the coiled-coil domain (CC) and a portion
of the talin homology domain (TH) was fused upstream the intracellular juxtamembrane region of ALK (Fig. 3C). The cDNA corresponding to the mRNA of HIP1–ALK in both patient and xenograft tumors were cloned and its complete
Copyright © 2014 by the International Association for the Study of Lung Cancer
289
Fang et al
Journal of Thoracic Oncology ® • Volume 9, Number 3, March 2014
FIGURE 3. Validation of ALK rearrangements by RT-PCR and direct sequencing. A, RT-PCR products corresponding to HIP1–ALK and EML4–ALK fusion genes, respectively, in both the primary patient (PA) and PDX tissues. M, DNA ladder. B, Direct sequencing of cDNA revealed fusion junctions of HIP1–ALK and EML4–ALK translocations in LU-01-0015 and LU-01-0319, respectively. The corresponding locations of the genes were indicated by color-coded bars. C, Schematic representation of HIP1–ALK protein. Regions corresponding to HIP1 and ALK were shown in blue and red, respectively. The fusion protein joins the tyrosine kinase domain of ALK to the TH domain of HIP1. RT-PCR, reverse transcription polymerase chain reaction; ALK, anaplastic lymphoma kinase; HIP1, huntingtin interacting protein 1; EML4, echinoderm microtubule-associated protein-like 4; PA, primary patient; PDX, patient-derived xenograft; TH, talin homology.
nucleotide sequence was determined by direct sequencing. The data revealed one nonsynonymous single-nucleotide polymorphism (SNP) in HIP1 portion, one synonymous SNP, three nonsynonymous SNPs in coding sequence, and one insertion and one SNP in 3′-untranslated region of ALK portion (Supplementary Figure 1A and 1B; Supplemental Digital Content 2, http://links.lww.com/JTO/A527). Compared with HIP1–ALK fusion of patient tumor, no secondary mutations were identified in the xenograft tumor. The cDNA contained an open reading frame coding for a protein of 1526 amino acids resulting from the in-frame fusion of the HIP1 N terminus (residues 1–963) with the ALK C terminus (residues 1058–1620) (Supplementary Figure 1C; Supplemental Digital Content 2, http://links.lww.com/JTO/A527). In LU-01-0319, EML4–ALK variants 3a and 3b were verified in both the primary tumor and xenograft (Fig. 3A and B, right panels). Analysis of sequence chromatograms suggested that EML4–ALK variant 3b was the dominant isoform in the xenograft tumor, as the dominant peaks corresponded
290
to variant 3b whereas the baseline noise peaks corresponded to variant 3a (Fig. 3B, lower right panel). Interestingly, expression level differences between variants 3a and 3b in the primary tumor were less remarkable in PDX tumors as the two peaks in a sequencing chromatogram of primary tumor overlapped with equal heights (Fig. 3B, upper right panel). These results suggest that EML4–ALK variant 3b was enriched during the establishment and/or serial passaging of PDX tumors.
PDX Tumors Retained Histopathological Characterizations of the Primary Tumors from Which They Were Derived ALK-rearranged NSCLC PDX models LU-01-0015 and LU-01-0319 were serially xenografted up to passage 3 in BALB/c nude mice. PDX tumors maintained a similar growth pattern on serial xenotransplantation (Fig. 4). Histopathology of primary and xenograft tumors was further evaluated by a pathologist. Both LU-01-0015 and LU-01-0319 were diagnosed as moderately differentiated squamous cell carcinoma,
Copyright © 2014 by the International Association for the Study of Lung Cancer
Journal of Thoracic Oncology ® • Volume 9, Number 3, March 2014
HIP1–ALK, a ALK Fusion Variant that Responds to Crizotinib
FIGURE 4. The histology of ALK-rearranged NSCLC patient (PA) and serially passaged PDX tumors and the growth curves of corresponding PDX tumors. The growth characteristics were summarized in the tables on the left. T250 and T500: the number of days required for a tumor to reach 250 mm3 and 500 mm3, respectively. T500-T250: the number of days required for a tumor to grow from 250 mm3 to 500 mm3. PDX, patient-derived xenograft; ALK, anaplastic lymphoma kinase; HIP1, huntingtin interacting protein 1; EML4, echinoderm microtubule-associated protein-like 4; NSCLC, non–small-cell lung cancer.
shown by the representative images of the primary tumors (Fig. 4, PA). Resection of the primary tumor from which LU-01-0015 was derived revealed a stage T3N1M0 NSCLC in a 62-year-old male patient with a 20-year smoking history. The primary tumor, which gave rise to LU-01-0319, was of stage T2N2M0 NSCLC in a 40-year-old male patient without a smoking history. The presence of keratin was observed in both primary and xenograft tumors of LU-01-0319 but barely existed in LU-01-0015. Xenografts at various passages truthfully recapitulated the heterogeneity of the primary tumors and retained histopathological features as squamous cell carcinoma (Fig. 4, P0–P3). These results showed that PDX tumors retained the histological features of the original tumors, exhibiting a concordance between xenografts and the parental patient tumors. To further confirm the histology of squamous cell carcinoma of both cases, immunohistochemistry was performed using a panel of differentiation markers, including CK5/6, 34βE12, P63, BerEP4, and MOC31. CK5/6, p63, and 34βE12 are the markers for squamous cell carcinoma. MOC31 and BerEP4 are expressed only in adenocarcinoma. The presence of CK5/6, p63, and 34βE12 expression and the absence of MOC31 and BerEP4 expression in both patient samples and
PDX tissues confirm that both cases belong to squamous cell carcinoma (Supplementary Figure 2, Supplemental Digital Content 3, http://links.lww.com/JTO/A528).
Responses of ALK-Rearranged NSCLC PDX Models to Crizotinib To elucidate the functional significance of ALK translocations, especially the novel HIP1–ALK fusion, in established PDX models, we then tested the responses of both models to crizotinib. An ALK wild-type model LU-01-0030 was randomly selected as a control (Supplementary Table 1, Supplemental Digital Content 1, http://links.lww.com/JTO/ A526). When the average tumor sizes reached 150 to 200 mm3, tumor-bearing animals were given crizotinib at 25 or 50 mg/kg (by mouth daily) or cisplatin at 3 mg/kg (by intraperitoneal, every week) or vehicle alone for 3 weeks. The results show that the average tumor volumes in crizotinib-treated groups were significantly less than those in the control groups in both LU-01-0015 and LU-01-0319, demonstrating that both models were sensitive to ALK inhibition (Fig. 5). Conversely, ALK wild-type model LU-01-0030 was resistant to crizotinib treatment. Differential responses to the standard-of-care drug, cisplatin, were observed in three models (Fig. 5).
Copyright © 2014 by the International Association for the Study of Lung Cancer
291
Journal of Thoracic Oncology ® • Volume 9, Number 3, March 2014
Fang et al
FIGURE 5. Antitumor activity of crizotinib in ALK wild-type, and HIP1–ALK- and EML4–ALK-rearranged NSCLC PDX models. Crizotinib (by mouth, daily) or cisplatin (by intraperitoneal every week) were administered in tumor-bearing mice for 3 weeks (n=6 or 7 in each group). Data were shown as mean ± standard error of the mean. In both ALK-rearranged models the average tumor volumes in the crizotinib- or cisplatin-treated groups at the endpoint are significantly less than the vehicle control group as determined using one-way analysis of variance. *p < 0.005, **p< 0.01, **p< 0.001. ALK, anaplastic lymphoma kinase; EML4, echinoderm microtubule-associated protein-like 4; HIP1, huntingtin interacting protein 1; PDX, patient-derived xenograft; NSCLC, non–small-cell lung cancer; WT, wild type.
DISCUSSION In the present study, we first screened ALK rearrangements in 99 established lung cancer PDX models by the NanoString ALK fusion assay (NanoString Technologies, Inc.). ALK rearrangements were identified in two NSCLC models (2%), which both also exhibited elevated ALK mRNA. Total RNA isolated from both xenograft tumors were subjected to RNA-Seq analyses. In addition to a previously known EML4– ALK fusion in one model, HIP1 was identified as a novel partner of ALK rearrangement in the other. RT-PCR and direct sequencing analyses of both primary tumor and xenograft elucidated that exon 28 of the HIP1 gene fused to a common breakpoint on exon 20 of the ALK gene. Furthermore, taking the advantage of available PDX models, we evaluated the effect of crizotinib in ALK wild-type, HIP1–ALK- and EML4– ALK-rearranged NSCLC PDX models, respectively. The results showed that, similar to the EML4–ALK-positive tumor, the HIP1–ALK-rearranged tumor responded sensitively to the treatment of crizotinib, whereas an ALK wild-type tumor did not respond to the treatment. Overall, our study not only identified HIP1 as a novel fusion partner of ALK in NSCLC, but also demonstrated the sensitivity of the HIP1–ALK rearrangement to crizotinib, warranting future clinical applications of crizotinib in HIP1–ALK-positive NSCLCs.
292
A subset of NSCLC harbors an inversion within chromosome 2 that gives rise to a transforming fusion gene, EML4– ALK, which encodes an activated protein tyrosine kinase. An EML4–ALK rearrangement in a lung adenocarcinoma was first identified in 2007.28 Such a fusion gene was also revealed in a lung adenocarcinoma cell line H2228.29 The discovery of the EML4–ALK fusion gene successfully promoted clinical trials in a subset of NSCLC harboring EML4–ALK fusion and led to the rapid approval of ALK inhibitor crizotinib by U.S. Food and Drug Administration. Multiple isoforms of EML4–ALK gene fusion have been discovered. In NSCLC, in addition to EML4, other genes are frequently fused to the ALK gene because of the presence of a common fragile locus within intron 19 of the ALK gene. Noteworthy examples include KIF5B–ALK, KLC1–ALK, and TFG–ALK in NSCLC.21 HIP1 belongs to the huntingtin interacting protein family (HIP1 and HIP1R in mammals and Sla2p in yeast), which plays a role in clathrin-mediated endocytosis and receptor trafficking.30–32 It consists of an N-terminally localized AP180 N-terminal homology domain (ANTH), a central coiled-coil domain (CC), and a talin homology domain (TH) at the C terminus (Fig. 3D).32,33 Fusion of HIP1 with PDGFβR was previously reported in a patient with chronic myelomonocytic leukemia
Copyright © 2014 by the International Association for the Study of Lung Cancer
Journal of Thoracic Oncology ® • Volume 9, Number 3, March 2014
with t(5;7)(q33; q11.2) translocation, in which the HIP1 gene located at 7q11.2 (i.e., exon 30) fused with the PDGFβR gene on 5q33.34 When expressed in the murine hematopoietic cell line, Ba/F3, HIP1–PDGFβR fusion protein becomes constitutively tyrosine phosphorylated and transforms the Ba/F3 cells to interleukin-3–independent growth, implicating its role in pathogenesis of hematopoietic malignancies.34,35 A subsequent study demonstrated that transgenic overexpression of HIP1 is associated with the development of lymphoid neoplasms and, therefore, HIP1 plays an important role in the transformation of lymphoid cells.32 In the present study, instead of exon 30, exon 28 of the HIP1 gene fused with exon 20 of the ALK gene in an NSCLC patient. These findings indicate that different isoforms of HIP1–ALK may exist. Analysis of novel HIP1–ALK translocations should contribute to our understanding of the pathogenesis of a subset of NSCLC. Although the molecular mechanism of transformation of HIP1–ALK translocations in lung carcinogenesis is still under investigation, the elevated ALK mRNA and in vivo efficacy caused by ALK inhibition suggest that fusion of HIP1 with ALK in NSCLC results in constitutive activation of ALK and triggers oncogenic cascade, as does the EML4–ALK fusion gene. HIP1–ALK cDNAs were obtained from total RNA isolated from both patient and PDX (P3) tissues, and then analyzed to elucidate the full coding sequence of the fusion gene. Interestingly, both sequences obtained from the patient sample and PDX tissue are identical, suggesting that no secondary mutation occurs after xenotransplantation (Supplementary Figure 1A–C; Supplemental Digital Content 2, http://links.lww.com/JTO/A527). In light of the facts that ALK rearragement predominantly exist in lung adenocarcinoma,36–39 it is interesting to note that, in our study, both cases of ALK rearrangements were squamous cell carcinoma. Apparently, the presence of EML4–ALK rearrangements has been described in various histological types of NSCLC, including squamous cell carcinoma,3,5,29 adenosquamous cancer,36 and adenocarcinoma with additional squamous or sarcomatoid differentiation.40 Therefore, although the frequency of ALK translocations seems to be lower than lung adenocarcinoma, other histological types of lung cancer also harbor ALK rearrangements. Preferable propagation of squamous cell carcinoma on xenotransplantation most likely contributes to the discovery of ALK rearrangements in two cases of squamous cell carcinoma reported in this study. In fact, the similar numbers of squamous cell carcinoma and adenocarcinoma patient samples were implanted; however, among the 99 lung cancer PDX models generated, adenocarcinoma and squamous cell carcinoma comprised 15% and 69%, respectively. These results suggest that the success rate of squmaous cell carcinoma is approximately fourfold higher than that of adenocarcinoma at xenotransplation. Higher frequency of squamous cell carcinoma increases its chance of detection of ALK rearrangement in the selected population of PDX models. Among ALK fusion variants identified in NSCLC, including EML4–ALK, KIF5B–ALK, KLC1–ALK, TFG–ALK, and ALK–PTPN3, the transforming potentials of the EML4– ALK, KIF5B–ALK, and KLC1–ALK fusion genes have been shown previously, implicating their role as oncogenic drivers of lung carcinogenesis.3,18,19,28,41 The biological significance
HIP1–ALK, a ALK Fusion Variant that Responds to Crizotinib
of other ALK fusion variants requires future investigation. However, inhibition of the catalytic activity of the ALK gene is required for ALK inhibition. It has been predicted that ALK– PTPN3 does not respond to crizotinib treatment because of its lack of the kinase domain. In the present study, the response of the HIP1–ALKpositive tumor to the treatment of crizotinib in the PDX model provides evidence of its oncogenic activity. Therefore, unlike several other ALK rearrangements in NSCLC, our studies provide evidence of the biological and clinical significance of HIP1–ALK fusion. As is the case for EML4–ALK translocations, HIP1–ALK may serve as a predictive biomarker for crizotinib treatment because HIP1–ALK confers sensitivity to the ALK inhibitor crizotinib. Patients with this chromosomal translocation most likely would derive clinical benefit from specific ALK inhibition. The frequency of HIP1–ALK fusion is 1% in our cohort (n=99). The incidence of the fusion should be investigated in a larger cohort of NSCLC. Clearly, PDX models preserve the genetic and histological heterogeneity of clinical patient tumors. Three of the NSCLC models studied here showed differential responses to the chemotherapy drug cisplatin, representing heterogenic responses to standard-of-care therapy observed in the patient population in the clinic. These xenografts have recently emerged as an excellent preclinical model for drug testing. Established PDX models harboring either EML4–ALK or HIP1–ALK fusion genes offer a test system for evaluating the next generation of ALK inhibitors under active development. Furthermore, to support discovery efforts and to better understand the mechanisms for drug resistance, we have initiated the development of crizotinib-resistant models in both EML4– ALK- and HIP1–ALK-containing PDX models in mice.
ACKNOWLEDGMENTS
The authors are grateful to Dan Lu, Hong Qin, Yan Xue, and Xi Wang of WuXi AppTec for technical support of this work. This work was partially supported by the National Basic Research Program of China (973 Program) Grant No. 2012CB724500. Anaplastic lyphoma kinase–rearranged patient-derived xenograft models are available at WuXi AppTec for drug tests. REFERENCES 1. Lynch TJ, Bell DW, Sordella R, et al. Activating mutations in the epidermal growth factor receptor underlying responsiveness of non-small-cell lung cancer to gefitinib. N Engl J Med 2004;350:2129–2139. 2. Paez JG, Janne PA, Lee JC, et al. EGFR mutations in lung cancer: correlation with clinical response to gefitinib therapy. Science 2004;304:1497–1500. 3. Soda M, Choi YL, Enomoto M, et al. Identification of the transforming EML4-ALK fusion gene in non-small-cell lung cancer. Nature 2007;448:561–566. 4. Christensen JG, Zou HY, Arango ME, et al. Cytoreductive antitumor activity of PF-2341066, a novel inhibitor of anaplastic lymphoma kinase and c-Met, in experimental models of anaplastic large-cell lymphoma. Mol Cancer Ther 2007;6:3314–3322. 5. Kwak EL, Bang YJ, Camidge DR, et al. Anaplastic lymphoma kinase inhibition in non-small-cell lung cancer. N Engl J Med 2010;363:1693–1703. 6. Mano H. Non-solid oncogenes in solid tumors: EML4-ALK fusion genes in lung cancer. Cancer Sci 2008;99:2349–2355.
Copyright © 2014 by the International Association for the Study of Lung Cancer
293
Fang et al
Journal of Thoracic Oncology ® • Volume 9, Number 3, March 2014
7. Horn L, Pao W. EML4-ALK: honing in on a new target in non-small-cell lung cancer. J Clin Oncol 2009;27:4232–4235. 8. Morris SW, Kirstein MN, Valentine MB, et al. Fusion of a kinase gene, ALK, to a nucleolar protein gene, NPM, in non-Hodgkin’s lymphoma. Science 1994;263:1281–1284. 9. Shiota M, Fujimoto J, Semba T, et al. Hyperphosphorylation of a novel 80 kDa protein-tyrosine kinase similar to Ltk in a human Ki-1 lymphoma cell line, AMS3. Oncogene 1994;9:1567–1574. 10. Hernandez L, Pinyol M, Hernandez S, et al. TRK-fused gene (TFG) is a new partner of ALK in anaplastic large cell lymphoma producing two structurally different TFG-ALK translocations. Blood 1999;94:3265–3268. 11. Touriol C, Greenland C, Lamant L, et al. Further demonstration of the diversity of chromosomal changes involving 2p23 in ALK-positive lymphoma: 2 cases expressing ALK kinase fused to CLTCL (clathrin chain polypeptide-like). Blood 2000;95:3204–3207. 12. Ma Z, Cools J, Marynen P, et al. Inv(2)(p23q35) in anaplastic large-cell lymphoma induces constitutive anaplastic lymphoma kinase (ALK) tyrosine kinase activation by fusion to ATIC, an enzyme involved in purine nucleotide biosynthesis. Blood 2000;95:2144–2149. 13. Trinei M, Lanfrancone L, Campo E, et al. A new variant anaplastic lymphoma kinase (ALK)-fusion protein (ATIC-ALK) in a case of ALKpositive anaplastic large cell lymphoma. Cancer Res 2000;60:793–798. 14. Cools J, Wlodarska I, Somers R, et al. Identification of novel fusion partners of ALK, the anaplastic lymphoma kinase, in anaplastic large-cell lymphoma and inflammatory myofibroblastic tumor. Gen Chrom Cancer 2002;34:354–362. 15. Tort F, Campo E, Pohlman B, et al. Heterogeneity of genomic breakpoints in MSN-ALK translocations in anaplastic large cell lymphoma. Hum Pathol 2004;35:1038–1041. 16. Kinoshita Y, Tajiri T, Ieiri S, et al. A case of an inflammatory myofibroblastic tumor in the lung which expressed TPM3-ALK gene fusion. Pediatr Surg Int 2007;23:595–599. 17. Armstrong F, Lamant L, Hieblot C, et al. TPM3-ALK expression induces changes in cytoskeleton organisation and confers higher metastatic capacities than other ALK fusion proteins. Eur J Cancer 2007;43:640–646. 18. Takeuchi K, Choi YL, Togashi Y, et al. KIF5B-ALK, a novel fusion oncokinase identified by an immunohistochemistry-based diagnostic system for ALK-positive lung cancer. Clin Cancer Res 2009;15:3143–3149. 19. Togashi Y, Soda M, Sakata S, et al. KLC1-ALK: a novel fusion in lung cancer identified using a formalin-fixed paraffin-embedded tissue only. PLoS One 2012;7:e31323. 20. Jung Y, Kim P, Keum J, et al. Discovery of ALK-PTPN3 gene fusion from human non-small cell lung carcinoma cell line using next generation RNA sequencing. Genes Chromosomes Cancer 2012;51:590–597. 21. Ou SH, Bartlett CH, Mino-Kenudson M, et al. Crizotinib for the treatment of ALK-rearranged non-small cell lung cancer: a success story to usher in the second decade of molecular targeted therapy in oncology. Oncologist 2012;17:1351–1375. 22. Lira ME, Kim TM, Huang D, et al. Multiplexed gene expression and fusion transcript analysis to detect ALK fusions in lung cancer. J Mol Diagn 2013;15:51–61. 23. Wong DW, Leung EL, So KK, et al. The EML4-ALK fusion gene is involved in various histologic types of lung cancers from nonsmokers with wild-type EGFR and KRAS. Cancer 2009;115:1723–1733.
294
24. Conway T, Wazny J, Bromage A, et al. Xenome—a tool for classifying reads from xenograft samples. Bioinformatics (Oxford, England) 2012;28:i172–178. 25. Kim D, Pertea G, Trapnell C, et al. TopHat2: accurate alignment of transcriptomes in the presence of insertions, deletions and gene fusions. Gen Biol 2013;14:R36. 26. Trapnell C, Williams BA, Pertea G, et al. Transcript assembly and quantification by RNA-Seq reveals unannotated transcripts and isoform switching during cell differentiation. Natur Biotechnol 2010;28:511–515. 27. Kim D, Salzberg SL. TopHat-Fusion: an algorithm for discovery of novel fusion transcripts. Gen Biol 2011;12:R72. 28. Choi YL, Takeuchi K, Soda M, et al. Identification of novel isoforms of the EML4-ALK transforming gene in non-small cell lung cancer. Cancer Res 2008;68:4971–4976. 29. Rikova K, Guo A, Zeng Q, et al. Global survey of phospho tyrosine signaling identifies oncogenic kinases in lung cancer. Cell 2007;131:1190–1203. 30. Waelter S, Scherzinger E, Hasenbank R, et al. The huntingtin interacting protein HIP1 is a clathrin and alpha-adaptin-binding protein involved in receptor-mediated endocytosis. Hum Mol Genet 2001;10:1807–1817. 31. Wilbur JD, Chen CY, Manalo V, et al. Actin binding by Hip1 (huntingtininteracting protein 1) and Hip1R (Hip1-related protein) is regulated by clathrin light chain. J Biol Chem 2008;283:32870–32879. 32. Bradley SV, Holland EC, Liu GY, et al. Huntingtin interacting protein 1 is a novel brain tumor marker that associates with epidermal growth factor receptor. Cancer Res 2007;67:3609–3615. 33. Gottfried I, Ehrlich M, Ashery U. The Sla2p/HIP1/HIP1R family: similar structure, similar function in endocytosis? Biochem Soc Trans 2010;38:187–191. 34. Ross TS, Bernard OA, Berger R, et al. Fusion of huntingtin interacting protein 1 to platelet-derived growth factor beta receptor (PDGFbetaR) in chronic myelomonocytic leukemia with t(5;7)(q33;q11.2). Blood 1998;91:4419–4426. 35. Ross TS, Gilliland DG. Transforming properties of the huntingtin interacting protein 1/ platelet-derived growth factor beta receptor fusion protein. J Biol Chem 1999;274:22328–22336. 36. Shaw AT, Yeap BY, Mino-Kenudson M, et al. Clinical features and outcome of patients with non-small-cell lung cancer who harbor EML4ALK. J Clin Oncol 2009;27:4247–4253. 37. Koivunen JP, Mermel C, Zejnullahu K, et al. EML4-ALK fusion gene and efficacy of an ALK kinase inhibitor in lung cancer. Clin Cancer Res 2008;14:4275–4283. 38. Takeuchi K, Choi YL, Soda M, et al. Multiplex reverse transcriptionPCR screening for EML4-ALK fusion transcripts. Clin Cancer Res 2008;14:6618–6624. 39. Takahashi T, Sonobe M, Kobayashi M, et al. Clinicopathologic features of non-small-cell lung cancer with EML4-ALK fusion gene. Ann Surg Oncol 2010;17:889–897. 40. Yoshida A, Tsuta K, Nakamura H, et al. Comprehensive histologic analysis of ALK-rearranged lung carcinomas. Am J Surg Pathol 2011;35:1226–1234. 41. Wong DW, Leung EL, Wong SK, et al. A novel KIF5B-ALK variant in nonsmall cell lung cancer. Cancer 2011;117:2709–2718.
Copyright © 2014 by the International Association for the Study of Lung Cancer
HIP1–ALK, A Novel ALK Fusion Variant that Responds to Crizotinib Douglas D. Fang, PhD,* Bin Zhang, PhD,* Qingyang Gu, PhD,* Maruja Lira, BS,‡ Qiang Xu, PhD,† Hongye Sun, PhD,† Maoxiang Qian, PhD,† Weiqi Sheng, MD, PhD,§ Mark Ozeck, BS,‡ Zhenxiong Wang, PhD,‡ Cathy Zhang, MS,‡ Xinsheng Chen, MS,* Kevin X. Chen, PhD,* Jian Li, PhD,* Shu-Hui Chen, PhD,* James Christensen, PhD,‡ Mao Mao, MD, PhD,‡ and Chi-Chung Chan, PhD* Introduction: The aim of this study was to identify anaplastic lymphoma kinase (ALK) rearrangements in lung cancer patient-derived xenograft (PDX) models and to explore their responses to crizotinib. Methods: Screening of 99 lung cancer PDX models by the NanoString ALK fusion assay identified two ALK-rearranged non– small-cell lung cancer (NSCLC) tumors, including one harboring a previously known echinoderm microtubule-associated protein-like 4 (EML4)–ALK fusion and another containing an unknown ALK fusion variant. Expression array, RNA-Seq, reverse transcription polymerase chain reaction, and direct sequencing were then conducted to confirm the rearrangements and to identify the novel fusion partner in the xenograft and/or the primary patient tumor. Finally, pharmacological studies were performed in PDX models to evaluate their responses to ALK inhibitor crizotinib. Results: Two ALK-rearranged NSCLC PDX models were identified: one carried a well-known EML4–ALK variant 3a/b and the other harbored a novel huntingtin interacting protein 1 (HIP1)–ALK fusion gene. Exon 28 of the HIP1 gene located on chromosome 7 was fused to exon 20 of the ALK gene located on chromosome 2. Both cases were clinically diagnosed as squamous cell carcinoma. Compared with the other lung cancer PDX models, both ALK-rearranged models displayed elevated ALK mRNA expression. Furthermore, in vivo efficacy studies demonstrated that, similar to the EML4–ALK-positive model, the HIP1–ALKcontaining PDX model was sensitive to treatment with crizotinib. Conclusions: Discovery of HIP1 as a fusion partner of ALK in NSCLC is a novel finding. In addition, the HIP1–ALK-rearranged *Discovery Services, †Genomics Center, WuXi AppTec Co., Ltd., Shanghai, China; ‡Pfizer Oncology Research Unit, San Diego, California; and §Department of Pathology, Shanghai Cancer Center, Fudan University, Shanghai, China. Drs. Zhang, and Gu contributed equally to this work. Drs. Mao and Chan contributed equally as senior authors. Disclosure: Several authors, as indicated above, are currently employees of Pfizer, at which crizotinib was invented, developed, and manufactured. The other authors declare no conflict of interest. Address for correspondence: Chi-Chung Chan, PhD, WuXi AppTec Co., Ltd., 288 Fute Zhong Road, Waigaoqiao Free Trade Zone, Shanghai, China 200131. E-mail: [email protected]; or Mao Mao, MD, PhD, Pfizer Oncology Unit, 10777 Science Center Drive, San Diego, CA 92121. E-mail: [email protected]. Copyright © 2014 by the International Association for the Study of Lung Cancer ISSN: 1556-0864/14/0903-0285
tumor is sensitive to treatment with crizotinib in vivo, implicating HIP1–ALK as an oncogenic driver of lung tumorigenesis. Collectively, our results indicate that HIP1–ALK-positive NSCLC may benefit from clinical applications of crizotinib. Key Words: Huntingtin interacting protein 1, anaplastic lymphoma kinase, HIP1–ALK, Echinoderm microtubule-associated proteinlike 4, EML4–ALK, Non–small-cell lung cancer, Crizotinib (PF02341066), Patient-derived xenograft (PDX) tumor model. (J Thorac Oncol. 2014;9: 285–294)
T
yrosine kinases have been widely recognized as attractive targets for molecularly targeted cancer therapy. Over the last decade a greater understanding of the molecular characteristics of non–small-cell lung cancer (NSCLC) has led to the development of effective targeted therapies for personalized medicine in a subset of patients. Targeting activating epidermal growth factor receptor (EGFR) mutations was the first breakthrough. Activation by mutation in EGFR in NSCLC correlated with clinical response to EGFR tyrosine kinase inhibitors.1,2 Afterward, in 2007, echinoderm microtubule-associated protein-like 4 (EML4)–ALK fusion-type tyrosine kinase was identified as an oncoprotein in 4% to 5% of NSCLC cases.3 Clinical trials were soon initiated to investigate the role of crizotinib (PF-02341066), an oral selective inhibitor of ALK and hepatocyte growth factor receptor (c-MET) kinases,4 for the treatment of EML4–ALK-positive NSCLC.5 On the basis of response rates reported in phase 1 and 2 clinical trials, crizotinib received accelerated approval by U.S. Food and Drug Administration in August 2011 for the treatment of patients with locally advanced or metastatic NSCLC that is ALK-positive as detected by the ALK fluorescence in situ hybridization test. EML4–ALK is generated as a result of a small inversion within the short arm of human chromosome 2.3,6,7 EML4–ALK undergoes constitutive dimerization through interaction between the coiled-coil domain within the EML4 region of each monomer, thereby activating ALK and generating oncogenic activity. Besides EML4, other fusion partners of ALK have been reported in a variety of cancer types (primarily in hematopoietic malignancies), including nucleophosmin (NPM),8,9 TRK-fused gene (TFG),10 clathrin heavy chain (Hc) (CLTC),11
Journal of Thoracic Oncology ® • Volume 9, Number 3, March 2014
285
Journal of Thoracic Oncology ® • Volume 9, Number 3, March 2014
Fang et al
5-aminoimidazole-4-carboxamide ribonucleotide formyltransferase/IMP cyclohydrolase (ATIC),12,13 cysteinyl-tRNA synthetase (CARS),14 moesin (MSN),15 tropomyosin 3 (TPM3),16,17 kinesin family member 5B (KIF5B),18 kinesin light chain 1 (KLC1),19 and tyrosine-protein phosphatase non-receptor type 3 (PTPN3).20 To date, only EML4–ALK, TFG–ALK, KIF5B–ALK, KLC1–ALK, and ALK–PTPN3 have been identified in lung cancer.21 Although the biological and clinical significance of these variants requires further investigation, it seems that not all of the above mentioned ALK rearrangements respond to crizotinib. For example, an ALK– PTPN3 variant may not respond to crizotinib, as the ALK kinase domain is absent.21 In addition, other, not-yet-characterized fusions may also exist in solid tumors, including lung cancer.6 We report here a novel ALK fusion associated with NSCLC involving the HIP1 gene and demonstrate the sensitivity of a HIP1–ALK-rearranged tumor to treatment with the ALK inhibitor crizotinib.
MATERIALS AND METHODS Establishment of NSCLC PDX Models In compliance with the protocol approved by the Institutional Review Board of Cancer Institute and Hospital of Chinese Academy of Medical Sciences and with the subject’s informed consent, primary tumor tissues were collected for PDX establishment. In brief, surgically resected primary tumor tissues from patients (designated as PA) were collected, trimmed, cut into 20- to 30-mm3 fragments and immediately implanted subcutaneously by using an 18-gauge trocar in the fore and/or hind bilateral flanks of 6- to 8-week-old female BALB/c nude mice (Shanghai SLAC Laboratory Animal Co., Ltd., Shanghai, China). Once the first generation of xenografts (designed as P0) was established and tumor sizes reached 500 to 800 mm3, serial implantations in BALB/c nude mice were performed to expand the xenograft tumors (i.e., P1, P2, P3, and beyond). Tumor size was measured periodically using a digital caliper (Cal Pro, Sylvac, Switzerland). Tumor volume was calculated as 0.5 × length × width.2 All procedures and protocols were approved by the Institutional Animal Care and Use Committee of WuXi AppTec.
Histology Pieces of patient samples or PDX tissues at each passage were fixed in 10% buffered formalin for 24 to 48 hours and then paraffin-embedded. Five μm sections were cut and stained with hematoxylin and eosin. Histopathology was reviewed by a pathologist (WS).
Total RNA Isolation Total RNA was isolated from the primary patient tumors and xenografts at P3 (LU-01-0015) or P2 (LU-01-0319) mouse models by using an RNeasy protect Mini Kit (Qiagen, Valencia, CA).
NanoString Assay for ALK Fusions The method for the detection of ALK fusions and expressions using NanoString’s nCounter technology (NanoString Technologies, Inc., Seattle, WA) has been described previously.22 In brief, after incubation of total RNA with nCounter probe sets, the samples were transferred to the nCounter Prep Station (NanoString Technologies, Inc.), where excess probes were
286
removed and probe/target complexes were aligned and immobilized in the nCounter Cartridge (NanoString Technologies, Inc.). Cartridges containing immobilized and aligned reporter complex were subsequently imaged on an nCounter Digital Analyzer (NanoString Technologies, Inc.) set at 1155 fields of view. Reporter counts were collected using NanoString’s nSolver analysis software version 1. The raw data were normalized and analyzed in two steps as previously described.22 A single-tube, multiplexed assay system was designed to simultaneously detect ALK fusion transcripts and measure the expression levels of several ALK exons flanking the fusion breakpoint. For fusion detection, ALK fusion partner EML4-, KIF5B-, or TFG-specific 5′ capture probes and ALK-specific 3′ reporter probes were designed to hybridize to ALK fusion partner and ALK at the fusion junction, respectively. The capture probes were able to detect all major variants of EML4-ALK fusions (variants 1, 2, 3a, 5, and “V5”21,23) and non–EML4–ALK fusions, including KIF5B–ALK and TFG– ALK. For ALK gene expression, probe sets across the entire ALK transcript, including four probe sets designated as ALK 5 ′-1 to 5 ′-4 located upstream of the intron 19 fusion breakpoint and four probe sets ALK 3 ′-1 to 3 ′-4 located downstream of the fusion breakpoint, were designed to measure the expression levels of ALK exons flanking the fusion breakpoint. For fusion detection, a reporter count of 60 was designated as the background threshold level, and the fusion was called present if the normalized reporter count was higher than background threshold. For ALK gene expression, the ALK 3′ overexpression score (i.e., ALK3 ′/5′ ratio) was defined as follows: ALK3′/5′ = E3/max(A5, B), where E3 is the geometric mean of ALK 3′ probe expression, A5 is the average of the ALK 5′ probe expression, and B is the background threshold. The fusion was called present if the ALK 3′ overexpression score was greater than two.
Microarray for ALK Expression Total RNA was amplified and fragmented using a GeneChip 3′ IVT Expression Kit (Affymetrix, Santa Clara, CA). Then the samples were hybridized onto a GeneChip PrimeView Human Gene Expression Array or GeneChip Human Genome U133 Plus 2.0 Array (Affymetrix) and scanned on a Affymetrix GeneChip Scanner 3000 7G (Affymetrix). Data were subject to in-house bioinformatics analysis. In brief, the raw CEL data of the PrimeView and U133 Plus 2.0 Arrays were processed on an Expression Console (version 1.1, Affymetrix). Signal intensities of both arrays were normalized by the Robust Multiarray Average normalization approach.
Paired-End RNA-Seq Analysis RNA-Seq library was generated using the Illumina TruSeq RNA Sample Prep Kit v2 (Illumina, San Diego, CA) and sequencing was performed on the Illumina HiSeq 2000 platform (Illumina). Human sequence reads were isolated from the mixture of reads arising from the mouse host and reads arising from xenograft tumors by using a specific tool Xenome.24 The purified reads were then mapped to the Ensembl GRCh37.62 B (hg19) reference genome using an RNA sequence aligner TopHat 2,25 which can align reads across splice junctions with or without relying on gene annotation. On the basis of the
Copyright © 2014 by the International Association for the Study of Lung Cancer
Journal of Thoracic Oncology ® • Volume 9, Number 3, March 2014
aligned result, transcript assembly and abundance estimation were performed with the software Cufflinks, a well-established transcript assembler,26 and fusion transcripts were discovered with the enhanced TopHat-Fusion module.27
Reverse Transcription Polymerase Chain Reaction and Direct Sequencing
Total RNA (1 μg) was reversely transcribed into single-stranded cDNAs using High Capacity cDNA Reverse Transcription Kits (Life Technologies, Carlsbad, CA) following the manufacturer’s instructions. One microliter of cDNA was used for a subsequent 25-μl polymerase chain reaction (PCR) amplification using a Premix PrimeSTAR HS PCR amplification kit (Takara Bio, Otsu, Shiga, Japan). To detect fusion transcript, we designed forward- and reverseprimer combinations targeting the sequences flanking the breakpoint detected by RNA-Seq. For HIP1–ALK fusion, the forward primer (5′-CGGCCAGAATAGAGGGTACA-3′) targeting the 5′ partner gene HIP1 and reverse primer (5′CACCTGGCCTTCATACACCT-3′) targeting the 3′ partner ALK were used. For EML4–ALK fusion, the previously described forward primer (5′-TACCAGTGCTGTCTCAATTGCAGG-3 ′) and reverse primer (5′-TCTTGCCAGCAAAGCAGTAGTTGG-3′) were used to detect EML4–ALK variant 3a/3b28. The PCR products were sequenced by conventional direct sequencing (BioSune Biotechnology (Shanghai) Co., Ltd., Shanghai, China) by using an ABI 3730 DNA Sequencer (Applied Biosystems, Foster City, CA). The results were compared with the reference sequence to validate the fusion transcripts.
In Vivo Efficacy Study Crizotinib [(R)-3-[1-(2,6-dichloro-3-fluoro-phenyl)ethoxy]-5-(1-piperidin4-yl-1H-pyrazol-4-yl)-pyridin-2-ylamine] was synthesized by WuXi AppTec (Shanghai, China) and formulated in distilled water for in vivo study. Injectable cisplatin for clinical use (10 mg/vial) was purchased from Qilu Pharmaceutical Co., Ltd (Jinan, Shandong, China). Tumor tissues were cut into small fragments of approximately 30 mm3 under sterile conditions. BALB/c nude mice were implanted subcutaneously with a tumor fragment by using a trocar. When the average tumor size reached 150 to 200 mm3, tumor size–matched mice were randomly assigned to four groups with six or seven mice in each group. The tumor-bearing mice were given ALK inhibitor crizotinib at 25 and 50 mg/kg (by mouth daily), cisplatin at 3 mg/kg (by intraperitoneal, every week), or vehicle alone for 3 weeks. Tumor volumes and body weights were measured using calipers twice a week. The difference in tumor volumes between treatment groups was analyzed for significance using one-way analysis of variance followed by Dunnett’s test. The p value less than 0.05 was considered to be statistically significant.
RESULTS In the past 2 years, we established more than 150 transplantable lung cancer PDX models from Chinese lung cancer patients. The first 99 of the 150 models, consisting of 96 with NSCLC, were used in this study, and their clinical information is summarized in Supplementary Table 1 (Supplemental Digital Content 1, http://links.lww.com/JTO/A526).
HIP1–ALK, a ALK Fusion Variant that Responds to Crizotinib
Detection of ALK Rearrangement in Lung Cancer PDX Models by Nanostring Fusion Assay To identify ALK rearrangements, RNA samples isolated from xenograft tissues of 99 models were analyzed by a combination of ALK 3′ overexpression- and ALK fusion-specific assays using NanoString’s nCounter technology. Because ALK is generally not expressed in normal adult tissues, ALK 3′ overexpression in lung cancer was indicative of an ALK rearrangement. In the ALK 3′ overexpression assay, we found that two NSCLC models, LU-01-0015 and LU-01-0319, exhibited ALK 3′ overexpression as shown in Figure 1A with an ALK 3′ overexpression score (ALK3 ′/5′ ratio) of 6.19 and 7.28, respectively (Fig. 1B). These results show that ALK is overexpressed in both NSCLC models, implicating ALK rearrangements. ALK fusion partner–specific 5′ capture probes and ALKspecific 3′ reporter probes were designed to detect the presence of preselected ALK fusion partners as previously described.22 Most ALK fusion variants share the same 3′ portion of ALK starting with exon 20. Thus, a common reporter probe (designated as ALK exon 20), paired with capture probes targeting different ALK fusion partners and variants, was able to detect the preselected fusion transcripts containing ALK exon 20 sequences, including all major variants of EML4–ALK fusions (variants 1, 2, 3a, 5, and V5) and KIF5B–ALK and TFG–ALK rearrangements. A reporter count of 60 was designated as the background threshold level, and the fusion was called present if the normalized reporter count was higher than background threshold. Among two models, LU-01-0319 showed higher reporter counts than the fusion reporter threshold, suggesting that it harbored the preselected fusion (Fig. 1C). Conversely, the reporter counts detected by the ALK exon 20 reporter were below the threshold in LU-01-0015. The data indicate that LU-01-0015 likely harbored a novel ALK fusion variant, which was not detectable by using the probe sets specific for known fusion genes indicated above.
Confirmation of ALK Overexpression in LU01-0015 and LU-01-0319 by Microarray The gene expression profiles of 29 lung cancer PDX models, including LU-01-0015 and LU-01-0319, were analyzed by a GeneChip PrimeView Human Gene Expression Array (Affymetrix; 7 models) or GeneChip Human Genome U133 Plus 2.0 Array (Affymetrix; 22 models). Analyses of ALK expression revealed that ALK mRNA levels in both LU-010015 and LU-01-0319 were markedly higher than in the other PDX models (Fig. 1D). These results provide further evidence for the presence of ALK rearrangements in both models.
Identification of HIP1–ALK by RNA-Seq To identify ALK fusion variants in LU-01-0015 and LU-01-0319, we used massively parallel, paired-end sequencing of expressed transcripts (RNA-Seq) to detect gene fusions as described in Materials and Methods. The RNA-Seq produced approximately 80.9 million and 105.5 million paired reads (100 base pair/read) for LU-01-0015 and LU-01-0319, respectively. Data analyses revealed that the ALK fusion gene in LU-010319 was a known EML4–ALK variant 3a/b, which was created by the breakage and intrachromosomal rearrangement of
Copyright © 2014 by the International Association for the Study of Lung Cancer
287
Fang et al
Journal of Thoracic Oncology ® • Volume 9, Number 3, March 2014
FIGURE 1. Detection of ALK rearrangements and overexpression of the ALK gene in two NSCLC PDX models. A, Quantification of ALK 5′ and 3′ expression levels using the NanoString ALK fusion assay. B, ALK 3′ overexpression in both models as measured by ALK 3′/5′ score in the fusion assay. A background threshold of twofold was denoted by a horizontal dash line (B). C, ALK exon 20 reporter counts detected in NanoString assay. A background threshold of 60 reporter counts was denoted by a horizontal dash line (B). D, Expression levels of ALK in NSCLC PDX models determined by a GeneChip PrimeView human gene expression array or a GeneChip Human Genome U133 Plus 2.0 array in a panel of 22 and seven models, respectively. ALK, anaplastic lymphoma kinase; PDX, patient–derived xenograft; NSCLC, non–small-cell lung cancer.
chromosome 2 (Fig. 2A and B).28 Of special interest is the identification of huntingtin interacting protein 1 (HIP1) as a novel fusion partner of ALK in the second model LU-01-0015 (Fig. 2A and B). HIP1–ALK rearrangement was formed by breakage and rejoining of chromosomes 7 and 2. Sixteen spanning reads and 2 spanning mate pairs were aligned to the cDNA breakpoints of HIP1 and ALK genes, which revealed a fusion product formed by exon 28 of HIP1 fused to exon 20 of ALK (Fig. 2B).
288
Validation of ALK Fusions in Both Primary Tumors and Xenografts by RT-PCR and Direct Sequencing To further confirm the fusion transcripts, RT-PCR was conducted by using forward- and reverse-primer combinations pairing the sequences flanking the breakpoint detected by RNA-Seq. Subsequent sequencing of cDNA confirmed the presence of HIP1–ALK rearrangement in both the
Copyright © 2014 by the International Association for the Study of Lung Cancer
Journal of Thoracic Oncology ® • Volume 9, Number 3, March 2014
HIP1–ALK, a ALK Fusion Variant that Responds to Crizotinib
FIGURE 2. Identification of a novel HIP1–ALK fusion by RNA-Seq. A, Circos plots depicting ALK rearrangements in LU-01-0015 and LU-01-0319. Chromosomes were drawn to scale around the rim of the circle, and data are plotted on these coordinates. The blue arc and red arc represent intrachromosomal and interchromosomal translocation, respectively. B, Representative aligned reads were shown to map across the exon–exon fusion junction between HIP1 or EML4 and exon 20 of ALK in LU-010015 and LU-01-0319, respectively, as illustrated in the schematic diagram. In LU-01-0319, the split-reads are shown aligning on the breakpoints of EML4–ALK variant 3a (A) and 3b (B). ALK, anaplastic lymphoma kinase; HIP1, huntingtin interacting protein 1; EML4, echinoderm microtubule-associated protein-like 4.
primary tumor and xenograft of LU-01-0015 (Fig. 3A and B, left panels). In the predicted fusion protein, the majority of N-terminal of HIP1 encompassing the AP180 N-terminal homology (ANTH), the coiled-coil domain (CC) and a portion
of the talin homology domain (TH) was fused upstream the intracellular juxtamembrane region of ALK (Fig. 3C). The cDNA corresponding to the mRNA of HIP1–ALK in both patient and xenograft tumors were cloned and its complete
Copyright © 2014 by the International Association for the Study of Lung Cancer
289
Fang et al
Journal of Thoracic Oncology ® • Volume 9, Number 3, March 2014
FIGURE 3. Validation of ALK rearrangements by RT-PCR and direct sequencing. A, RT-PCR products corresponding to HIP1–ALK and EML4–ALK fusion genes, respectively, in both the primary patient (PA) and PDX tissues. M, DNA ladder. B, Direct sequencing of cDNA revealed fusion junctions of HIP1–ALK and EML4–ALK translocations in LU-01-0015 and LU-01-0319, respectively. The corresponding locations of the genes were indicated by color-coded bars. C, Schematic representation of HIP1–ALK protein. Regions corresponding to HIP1 and ALK were shown in blue and red, respectively. The fusion protein joins the tyrosine kinase domain of ALK to the TH domain of HIP1. RT-PCR, reverse transcription polymerase chain reaction; ALK, anaplastic lymphoma kinase; HIP1, huntingtin interacting protein 1; EML4, echinoderm microtubule-associated protein-like 4; PA, primary patient; PDX, patient-derived xenograft; TH, talin homology.
nucleotide sequence was determined by direct sequencing. The data revealed one nonsynonymous single-nucleotide polymorphism (SNP) in HIP1 portion, one synonymous SNP, three nonsynonymous SNPs in coding sequence, and one insertion and one SNP in 3′-untranslated region of ALK portion (Supplementary Figure 1A and 1B; Supplemental Digital Content 2, http://links.lww.com/JTO/A527). Compared with HIP1–ALK fusion of patient tumor, no secondary mutations were identified in the xenograft tumor. The cDNA contained an open reading frame coding for a protein of 1526 amino acids resulting from the in-frame fusion of the HIP1 N terminus (residues 1–963) with the ALK C terminus (residues 1058–1620) (Supplementary Figure 1C; Supplemental Digital Content 2, http://links.lww.com/JTO/A527). In LU-01-0319, EML4–ALK variants 3a and 3b were verified in both the primary tumor and xenograft (Fig. 3A and B, right panels). Analysis of sequence chromatograms suggested that EML4–ALK variant 3b was the dominant isoform in the xenograft tumor, as the dominant peaks corresponded
290
to variant 3b whereas the baseline noise peaks corresponded to variant 3a (Fig. 3B, lower right panel). Interestingly, expression level differences between variants 3a and 3b in the primary tumor were less remarkable in PDX tumors as the two peaks in a sequencing chromatogram of primary tumor overlapped with equal heights (Fig. 3B, upper right panel). These results suggest that EML4–ALK variant 3b was enriched during the establishment and/or serial passaging of PDX tumors.
PDX Tumors Retained Histopathological Characterizations of the Primary Tumors from Which They Were Derived ALK-rearranged NSCLC PDX models LU-01-0015 and LU-01-0319 were serially xenografted up to passage 3 in BALB/c nude mice. PDX tumors maintained a similar growth pattern on serial xenotransplantation (Fig. 4). Histopathology of primary and xenograft tumors was further evaluated by a pathologist. Both LU-01-0015 and LU-01-0319 were diagnosed as moderately differentiated squamous cell carcinoma,
Copyright © 2014 by the International Association for the Study of Lung Cancer
Journal of Thoracic Oncology ® • Volume 9, Number 3, March 2014
HIP1–ALK, a ALK Fusion Variant that Responds to Crizotinib
FIGURE 4. The histology of ALK-rearranged NSCLC patient (PA) and serially passaged PDX tumors and the growth curves of corresponding PDX tumors. The growth characteristics were summarized in the tables on the left. T250 and T500: the number of days required for a tumor to reach 250 mm3 and 500 mm3, respectively. T500-T250: the number of days required for a tumor to grow from 250 mm3 to 500 mm3. PDX, patient-derived xenograft; ALK, anaplastic lymphoma kinase; HIP1, huntingtin interacting protein 1; EML4, echinoderm microtubule-associated protein-like 4; NSCLC, non–small-cell lung cancer.
shown by the representative images of the primary tumors (Fig. 4, PA). Resection of the primary tumor from which LU-01-0015 was derived revealed a stage T3N1M0 NSCLC in a 62-year-old male patient with a 20-year smoking history. The primary tumor, which gave rise to LU-01-0319, was of stage T2N2M0 NSCLC in a 40-year-old male patient without a smoking history. The presence of keratin was observed in both primary and xenograft tumors of LU-01-0319 but barely existed in LU-01-0015. Xenografts at various passages truthfully recapitulated the heterogeneity of the primary tumors and retained histopathological features as squamous cell carcinoma (Fig. 4, P0–P3). These results showed that PDX tumors retained the histological features of the original tumors, exhibiting a concordance between xenografts and the parental patient tumors. To further confirm the histology of squamous cell carcinoma of both cases, immunohistochemistry was performed using a panel of differentiation markers, including CK5/6, 34βE12, P63, BerEP4, and MOC31. CK5/6, p63, and 34βE12 are the markers for squamous cell carcinoma. MOC31 and BerEP4 are expressed only in adenocarcinoma. The presence of CK5/6, p63, and 34βE12 expression and the absence of MOC31 and BerEP4 expression in both patient samples and
PDX tissues confirm that both cases belong to squamous cell carcinoma (Supplementary Figure 2, Supplemental Digital Content 3, http://links.lww.com/JTO/A528).
Responses of ALK-Rearranged NSCLC PDX Models to Crizotinib To elucidate the functional significance of ALK translocations, especially the novel HIP1–ALK fusion, in established PDX models, we then tested the responses of both models to crizotinib. An ALK wild-type model LU-01-0030 was randomly selected as a control (Supplementary Table 1, Supplemental Digital Content 1, http://links.lww.com/JTO/ A526). When the average tumor sizes reached 150 to 200 mm3, tumor-bearing animals were given crizotinib at 25 or 50 mg/kg (by mouth daily) or cisplatin at 3 mg/kg (by intraperitoneal, every week) or vehicle alone for 3 weeks. The results show that the average tumor volumes in crizotinib-treated groups were significantly less than those in the control groups in both LU-01-0015 and LU-01-0319, demonstrating that both models were sensitive to ALK inhibition (Fig. 5). Conversely, ALK wild-type model LU-01-0030 was resistant to crizotinib treatment. Differential responses to the standard-of-care drug, cisplatin, were observed in three models (Fig. 5).
Copyright © 2014 by the International Association for the Study of Lung Cancer
291
Journal of Thoracic Oncology ® • Volume 9, Number 3, March 2014
Fang et al
FIGURE 5. Antitumor activity of crizotinib in ALK wild-type, and HIP1–ALK- and EML4–ALK-rearranged NSCLC PDX models. Crizotinib (by mouth, daily) or cisplatin (by intraperitoneal every week) were administered in tumor-bearing mice for 3 weeks (n=6 or 7 in each group). Data were shown as mean ± standard error of the mean. In both ALK-rearranged models the average tumor volumes in the crizotinib- or cisplatin-treated groups at the endpoint are significantly less than the vehicle control group as determined using one-way analysis of variance. *p < 0.005, **p< 0.01, **p< 0.001. ALK, anaplastic lymphoma kinase; EML4, echinoderm microtubule-associated protein-like 4; HIP1, huntingtin interacting protein 1; PDX, patient-derived xenograft; NSCLC, non–small-cell lung cancer; WT, wild type.
DISCUSSION In the present study, we first screened ALK rearrangements in 99 established lung cancer PDX models by the NanoString ALK fusion assay (NanoString Technologies, Inc.). ALK rearrangements were identified in two NSCLC models (2%), which both also exhibited elevated ALK mRNA. Total RNA isolated from both xenograft tumors were subjected to RNA-Seq analyses. In addition to a previously known EML4– ALK fusion in one model, HIP1 was identified as a novel partner of ALK rearrangement in the other. RT-PCR and direct sequencing analyses of both primary tumor and xenograft elucidated that exon 28 of the HIP1 gene fused to a common breakpoint on exon 20 of the ALK gene. Furthermore, taking the advantage of available PDX models, we evaluated the effect of crizotinib in ALK wild-type, HIP1–ALK- and EML4– ALK-rearranged NSCLC PDX models, respectively. The results showed that, similar to the EML4–ALK-positive tumor, the HIP1–ALK-rearranged tumor responded sensitively to the treatment of crizotinib, whereas an ALK wild-type tumor did not respond to the treatment. Overall, our study not only identified HIP1 as a novel fusion partner of ALK in NSCLC, but also demonstrated the sensitivity of the HIP1–ALK rearrangement to crizotinib, warranting future clinical applications of crizotinib in HIP1–ALK-positive NSCLCs.
292
A subset of NSCLC harbors an inversion within chromosome 2 that gives rise to a transforming fusion gene, EML4– ALK, which encodes an activated protein tyrosine kinase. An EML4–ALK rearrangement in a lung adenocarcinoma was first identified in 2007.28 Such a fusion gene was also revealed in a lung adenocarcinoma cell line H2228.29 The discovery of the EML4–ALK fusion gene successfully promoted clinical trials in a subset of NSCLC harboring EML4–ALK fusion and led to the rapid approval of ALK inhibitor crizotinib by U.S. Food and Drug Administration. Multiple isoforms of EML4–ALK gene fusion have been discovered. In NSCLC, in addition to EML4, other genes are frequently fused to the ALK gene because of the presence of a common fragile locus within intron 19 of the ALK gene. Noteworthy examples include KIF5B–ALK, KLC1–ALK, and TFG–ALK in NSCLC.21 HIP1 belongs to the huntingtin interacting protein family (HIP1 and HIP1R in mammals and Sla2p in yeast), which plays a role in clathrin-mediated endocytosis and receptor trafficking.30–32 It consists of an N-terminally localized AP180 N-terminal homology domain (ANTH), a central coiled-coil domain (CC), and a talin homology domain (TH) at the C terminus (Fig. 3D).32,33 Fusion of HIP1 with PDGFβR was previously reported in a patient with chronic myelomonocytic leukemia
Copyright © 2014 by the International Association for the Study of Lung Cancer
Journal of Thoracic Oncology ® • Volume 9, Number 3, March 2014
with t(5;7)(q33; q11.2) translocation, in which the HIP1 gene located at 7q11.2 (i.e., exon 30) fused with the PDGFβR gene on 5q33.34 When expressed in the murine hematopoietic cell line, Ba/F3, HIP1–PDGFβR fusion protein becomes constitutively tyrosine phosphorylated and transforms the Ba/F3 cells to interleukin-3–independent growth, implicating its role in pathogenesis of hematopoietic malignancies.34,35 A subsequent study demonstrated that transgenic overexpression of HIP1 is associated with the development of lymphoid neoplasms and, therefore, HIP1 plays an important role in the transformation of lymphoid cells.32 In the present study, instead of exon 30, exon 28 of the HIP1 gene fused with exon 20 of the ALK gene in an NSCLC patient. These findings indicate that different isoforms of HIP1–ALK may exist. Analysis of novel HIP1–ALK translocations should contribute to our understanding of the pathogenesis of a subset of NSCLC. Although the molecular mechanism of transformation of HIP1–ALK translocations in lung carcinogenesis is still under investigation, the elevated ALK mRNA and in vivo efficacy caused by ALK inhibition suggest that fusion of HIP1 with ALK in NSCLC results in constitutive activation of ALK and triggers oncogenic cascade, as does the EML4–ALK fusion gene. HIP1–ALK cDNAs were obtained from total RNA isolated from both patient and PDX (P3) tissues, and then analyzed to elucidate the full coding sequence of the fusion gene. Interestingly, both sequences obtained from the patient sample and PDX tissue are identical, suggesting that no secondary mutation occurs after xenotransplantation (Supplementary Figure 1A–C; Supplemental Digital Content 2, http://links.lww.com/JTO/A527). In light of the facts that ALK rearragement predominantly exist in lung adenocarcinoma,36–39 it is interesting to note that, in our study, both cases of ALK rearrangements were squamous cell carcinoma. Apparently, the presence of EML4–ALK rearrangements has been described in various histological types of NSCLC, including squamous cell carcinoma,3,5,29 adenosquamous cancer,36 and adenocarcinoma with additional squamous or sarcomatoid differentiation.40 Therefore, although the frequency of ALK translocations seems to be lower than lung adenocarcinoma, other histological types of lung cancer also harbor ALK rearrangements. Preferable propagation of squamous cell carcinoma on xenotransplantation most likely contributes to the discovery of ALK rearrangements in two cases of squamous cell carcinoma reported in this study. In fact, the similar numbers of squamous cell carcinoma and adenocarcinoma patient samples were implanted; however, among the 99 lung cancer PDX models generated, adenocarcinoma and squamous cell carcinoma comprised 15% and 69%, respectively. These results suggest that the success rate of squmaous cell carcinoma is approximately fourfold higher than that of adenocarcinoma at xenotransplation. Higher frequency of squamous cell carcinoma increases its chance of detection of ALK rearrangement in the selected population of PDX models. Among ALK fusion variants identified in NSCLC, including EML4–ALK, KIF5B–ALK, KLC1–ALK, TFG–ALK, and ALK–PTPN3, the transforming potentials of the EML4– ALK, KIF5B–ALK, and KLC1–ALK fusion genes have been shown previously, implicating their role as oncogenic drivers of lung carcinogenesis.3,18,19,28,41 The biological significance
HIP1–ALK, a ALK Fusion Variant that Responds to Crizotinib
of other ALK fusion variants requires future investigation. However, inhibition of the catalytic activity of the ALK gene is required for ALK inhibition. It has been predicted that ALK– PTPN3 does not respond to crizotinib treatment because of its lack of the kinase domain. In the present study, the response of the HIP1–ALKpositive tumor to the treatment of crizotinib in the PDX model provides evidence of its oncogenic activity. Therefore, unlike several other ALK rearrangements in NSCLC, our studies provide evidence of the biological and clinical significance of HIP1–ALK fusion. As is the case for EML4–ALK translocations, HIP1–ALK may serve as a predictive biomarker for crizotinib treatment because HIP1–ALK confers sensitivity to the ALK inhibitor crizotinib. Patients with this chromosomal translocation most likely would derive clinical benefit from specific ALK inhibition. The frequency of HIP1–ALK fusion is 1% in our cohort (n=99). The incidence of the fusion should be investigated in a larger cohort of NSCLC. Clearly, PDX models preserve the genetic and histological heterogeneity of clinical patient tumors. Three of the NSCLC models studied here showed differential responses to the chemotherapy drug cisplatin, representing heterogenic responses to standard-of-care therapy observed in the patient population in the clinic. These xenografts have recently emerged as an excellent preclinical model for drug testing. Established PDX models harboring either EML4–ALK or HIP1–ALK fusion genes offer a test system for evaluating the next generation of ALK inhibitors under active development. Furthermore, to support discovery efforts and to better understand the mechanisms for drug resistance, we have initiated the development of crizotinib-resistant models in both EML4– ALK- and HIP1–ALK-containing PDX models in mice.
ACKNOWLEDGMENTS
The authors are grateful to Dan Lu, Hong Qin, Yan Xue, and Xi Wang of WuXi AppTec for technical support of this work. This work was partially supported by the National Basic Research Program of China (973 Program) Grant No. 2012CB724500. Anaplastic lyphoma kinase–rearranged patient-derived xenograft models are available at WuXi AppTec for drug tests. REFERENCES 1. Lynch TJ, Bell DW, Sordella R, et al. Activating mutations in the epidermal growth factor receptor underlying responsiveness of non-small-cell lung cancer to gefitinib. N Engl J Med 2004;350:2129–2139. 2. Paez JG, Janne PA, Lee JC, et al. EGFR mutations in lung cancer: correlation with clinical response to gefitinib therapy. Science 2004;304:1497–1500. 3. Soda M, Choi YL, Enomoto M, et al. Identification of the transforming EML4-ALK fusion gene in non-small-cell lung cancer. Nature 2007;448:561–566. 4. Christensen JG, Zou HY, Arango ME, et al. Cytoreductive antitumor activity of PF-2341066, a novel inhibitor of anaplastic lymphoma kinase and c-Met, in experimental models of anaplastic large-cell lymphoma. Mol Cancer Ther 2007;6:3314–3322. 5. Kwak EL, Bang YJ, Camidge DR, et al. Anaplastic lymphoma kinase inhibition in non-small-cell lung cancer. N Engl J Med 2010;363:1693–1703. 6. Mano H. Non-solid oncogenes in solid tumors: EML4-ALK fusion genes in lung cancer. Cancer Sci 2008;99:2349–2355.
Copyright © 2014 by the International Association for the Study of Lung Cancer
293
Fang et al
Journal of Thoracic Oncology ® • Volume 9, Number 3, March 2014
7. Horn L, Pao W. EML4-ALK: honing in on a new target in non-small-cell lung cancer. J Clin Oncol 2009;27:4232–4235. 8. Morris SW, Kirstein MN, Valentine MB, et al. Fusion of a kinase gene, ALK, to a nucleolar protein gene, NPM, in non-Hodgkin’s lymphoma. Science 1994;263:1281–1284. 9. Shiota M, Fujimoto J, Semba T, et al. Hyperphosphorylation of a novel 80 kDa protein-tyrosine kinase similar to Ltk in a human Ki-1 lymphoma cell line, AMS3. Oncogene 1994;9:1567–1574. 10. Hernandez L, Pinyol M, Hernandez S, et al. TRK-fused gene (TFG) is a new partner of ALK in anaplastic large cell lymphoma producing two structurally different TFG-ALK translocations. Blood 1999;94:3265–3268. 11. Touriol C, Greenland C, Lamant L, et al. Further demonstration of the diversity of chromosomal changes involving 2p23 in ALK-positive lymphoma: 2 cases expressing ALK kinase fused to CLTCL (clathrin chain polypeptide-like). Blood 2000;95:3204–3207. 12. Ma Z, Cools J, Marynen P, et al. Inv(2)(p23q35) in anaplastic large-cell lymphoma induces constitutive anaplastic lymphoma kinase (ALK) tyrosine kinase activation by fusion to ATIC, an enzyme involved in purine nucleotide biosynthesis. Blood 2000;95:2144–2149. 13. Trinei M, Lanfrancone L, Campo E, et al. A new variant anaplastic lymphoma kinase (ALK)-fusion protein (ATIC-ALK) in a case of ALKpositive anaplastic large cell lymphoma. Cancer Res 2000;60:793–798. 14. Cools J, Wlodarska I, Somers R, et al. Identification of novel fusion partners of ALK, the anaplastic lymphoma kinase, in anaplastic large-cell lymphoma and inflammatory myofibroblastic tumor. Gen Chrom Cancer 2002;34:354–362. 15. Tort F, Campo E, Pohlman B, et al. Heterogeneity of genomic breakpoints in MSN-ALK translocations in anaplastic large cell lymphoma. Hum Pathol 2004;35:1038–1041. 16. Kinoshita Y, Tajiri T, Ieiri S, et al. A case of an inflammatory myofibroblastic tumor in the lung which expressed TPM3-ALK gene fusion. Pediatr Surg Int 2007;23:595–599. 17. Armstrong F, Lamant L, Hieblot C, et al. TPM3-ALK expression induces changes in cytoskeleton organisation and confers higher metastatic capacities than other ALK fusion proteins. Eur J Cancer 2007;43:640–646. 18. Takeuchi K, Choi YL, Togashi Y, et al. KIF5B-ALK, a novel fusion oncokinase identified by an immunohistochemistry-based diagnostic system for ALK-positive lung cancer. Clin Cancer Res 2009;15:3143–3149. 19. Togashi Y, Soda M, Sakata S, et al. KLC1-ALK: a novel fusion in lung cancer identified using a formalin-fixed paraffin-embedded tissue only. PLoS One 2012;7:e31323. 20. Jung Y, Kim P, Keum J, et al. Discovery of ALK-PTPN3 gene fusion from human non-small cell lung carcinoma cell line using next generation RNA sequencing. Genes Chromosomes Cancer 2012;51:590–597. 21. Ou SH, Bartlett CH, Mino-Kenudson M, et al. Crizotinib for the treatment of ALK-rearranged non-small cell lung cancer: a success story to usher in the second decade of molecular targeted therapy in oncology. Oncologist 2012;17:1351–1375. 22. Lira ME, Kim TM, Huang D, et al. Multiplexed gene expression and fusion transcript analysis to detect ALK fusions in lung cancer. J Mol Diagn 2013;15:51–61. 23. Wong DW, Leung EL, So KK, et al. The EML4-ALK fusion gene is involved in various histologic types of lung cancers from nonsmokers with wild-type EGFR and KRAS. Cancer 2009;115:1723–1733.
294
24. Conway T, Wazny J, Bromage A, et al. Xenome—a tool for classifying reads from xenograft samples. Bioinformatics (Oxford, England) 2012;28:i172–178. 25. Kim D, Pertea G, Trapnell C, et al. TopHat2: accurate alignment of transcriptomes in the presence of insertions, deletions and gene fusions. Gen Biol 2013;14:R36. 26. Trapnell C, Williams BA, Pertea G, et al. Transcript assembly and quantification by RNA-Seq reveals unannotated transcripts and isoform switching during cell differentiation. Natur Biotechnol 2010;28:511–515. 27. Kim D, Salzberg SL. TopHat-Fusion: an algorithm for discovery of novel fusion transcripts. Gen Biol 2011;12:R72. 28. Choi YL, Takeuchi K, Soda M, et al. Identification of novel isoforms of the EML4-ALK transforming gene in non-small cell lung cancer. Cancer Res 2008;68:4971–4976. 29. Rikova K, Guo A, Zeng Q, et al. Global survey of phospho tyrosine signaling identifies oncogenic kinases in lung cancer. Cell 2007;131:1190–1203. 30. Waelter S, Scherzinger E, Hasenbank R, et al. The huntingtin interacting protein HIP1 is a clathrin and alpha-adaptin-binding protein involved in receptor-mediated endocytosis. Hum Mol Genet 2001;10:1807–1817. 31. Wilbur JD, Chen CY, Manalo V, et al. Actin binding by Hip1 (huntingtininteracting protein 1) and Hip1R (Hip1-related protein) is regulated by clathrin light chain. J Biol Chem 2008;283:32870–32879. 32. Bradley SV, Holland EC, Liu GY, et al. Huntingtin interacting protein 1 is a novel brain tumor marker that associates with epidermal growth factor receptor. Cancer Res 2007;67:3609–3615. 33. Gottfried I, Ehrlich M, Ashery U. The Sla2p/HIP1/HIP1R family: similar structure, similar function in endocytosis? Biochem Soc Trans 2010;38:187–191. 34. Ross TS, Bernard OA, Berger R, et al. Fusion of huntingtin interacting protein 1 to platelet-derived growth factor beta receptor (PDGFbetaR) in chronic myelomonocytic leukemia with t(5;7)(q33;q11.2). Blood 1998;91:4419–4426. 35. Ross TS, Gilliland DG. Transforming properties of the huntingtin interacting protein 1/ platelet-derived growth factor beta receptor fusion protein. J Biol Chem 1999;274:22328–22336. 36. Shaw AT, Yeap BY, Mino-Kenudson M, et al. Clinical features and outcome of patients with non-small-cell lung cancer who harbor EML4ALK. J Clin Oncol 2009;27:4247–4253. 37. Koivunen JP, Mermel C, Zejnullahu K, et al. EML4-ALK fusion gene and efficacy of an ALK kinase inhibitor in lung cancer. Clin Cancer Res 2008;14:4275–4283. 38. Takeuchi K, Choi YL, Soda M, et al. Multiplex reverse transcriptionPCR screening for EML4-ALK fusion transcripts. Clin Cancer Res 2008;14:6618–6624. 39. Takahashi T, Sonobe M, Kobayashi M, et al. Clinicopathologic features of non-small-cell lung cancer with EML4-ALK fusion gene. Ann Surg Oncol 2010;17:889–897. 40. Yoshida A, Tsuta K, Nakamura H, et al. Comprehensive histologic analysis of ALK-rearranged lung carcinomas. Am J Surg Pathol 2011;35:1226–1234. 41. Wong DW, Leung EL, Wong SK, et al. A novel KIF5B-ALK variant in nonsmall cell lung cancer. Cancer 2011;117:2709–2718.
Copyright © 2014 by the International Association for the Study of Lung Cancer
